In world wide web, a document is usually made up of multiple pages, each one of which has a unique URL address and links to each other by hyperlink pointers. Related documents are also connected together through hyperlinks. Thus, it is important for web information discovery to know the referral parent(s) of a given web page. However, due to the lack of back-pointers, such information is usually not available to web surfers. This reduces the effectiveness of web surfing and information discovery significantly. Major search engines such as AltaVista and HotBot try to solve this problem from the server side by maintaining the hyperlink structure of each web site registered with them. Although effective, the drawbacks of this approach are the huge effort to store and maintain this link database and the limitation to their registered members only. In this paper, we propose an alternative mechanism to locate the referral parent(s) for a given fragment of a web document (in terms of an URL ...